Sabiacea.} FLORA INDICA. 207 lares, extrorsae vel introrsso, longitudinaliter dehiscentes, valvis a con- nectivo solutis, quapropter antherse post deliiscentiam uniloculares flunk Discus Jiypogymts colunmrc hrevi insidens, 5-lobus, lobis carnosis cum petalis scpalisque alternantibus. Ocaria 2, rarissime 3, in axi subco- haerentia, biomlata; ovnla suturac veutrali insertn, superposita, inferius descendeiis, cauipylotropum, superius fere horizontale, suborthotropum. Styli 2, erecti, terminales, cylindrici, secus faciem ventralem subco- hserentes, scd facile separabiles. Stigmata simplicia, obtusiuscula. Carpella % vel abortu solitaria, drupacea, dorso gibbosa, intus stylo subpersistentc fare basilari rostrata. 'Endocardium lignosura, irregu- lariter rugosum. Semen solitarium, rcniforme, prope basin insertutn, oampylotropum. Testa coriacca, punctis coloratis notata. Endopleura crasshiscula, alba. Embryo exalbuminosus, radicula infera horizontal! rylimlrica, cotyledonibus ovalibus incurvis planiusculis cnrnosis.— Vruticcs scandentes foliosi, ramulis bad squamis gemma perskteutibm stipatis, foliis altcm'tn mtegerrimis e$stvpulatis, cum petiolo fraud articu- latis, floi'ibus axillaribn^ solitarils cymosis vel paniculatis, Mediocribus vel parvis, mridibits jlavis vel purpureis, plerumque cum- foliis nascentibus evoluiis, The gcrnis Sabia was first described by Colobrooko* in the year 1820, with a aomewhal erroneous generic Hiarai'ier, and a plate \\liidi neenratcly represents the habit ami general aypenraueu of the ]>liint, but is ancotnpanieil by a very imperfect figure of the ilowcr. In 1S-4, AValliehf published excellent descriptions'of two ad- ditional species, giving \\\ the same time a corrected generic character, and referring the genus to TurGbitit/uwftr. In 1825, Itluiue,:); unaware of what had previously been done, added another speeies under the generic name Memscosta, which he placed at the end of Me?rispei'mace(e. Endlioher and JMeisner, adopting "Wallich/s suggestion, placed Sabia at the end of AnaraTttiacea?. In 1842, Falconer§ published an excellent account of the genus, under tbo name of 7vV'#//'//, which he indicated as the type of a distinct Order, pointing out the resemblance of the fruit to Memsyer- macetp9 but not pronouncing definitely on its affinities. In 1831, Blumc,[| who had discovered the identity of^jfer'genua Jlle.mcosfa with XaAit/i eonxtituted the Natural Order Su&iticfff* the place of \vliieh he fixed in the immediate viciuify of Menisper- wacete; and in Ib53 MjerR*. adopted that Natural Order, taking the same view of its affinities. He has, however, fallen into an error in describing the ovules as soli- tary, and has overlooked the remarkable character of the opposition, of the jxitals and sepalrt. The structure of the genus Sabia is so remarkable, that its claims to form a dis- tinct Order are unquestionable; but, as in the case of many Orders of limited extent, the characters point in so many different directions that-it is not easy to determine the position which it ought to occupy in our systems. If the ovary of Saliacea be considered synearpous, the presence of a well-marked" hypogynous disc, and many other characters, would seem to indicate the Rhamnal alliance as that to which they are most nearly allied. Among its Orders, CAattMiaccte, which have n two-celled ovary, containing two collateral pendulous ovules in eae.h cell, a simple style, and exalbuminous seeds, appear to exhibit the greatest amount of resemblance to Sabiacea. There are, however, many obvious differences, such as the structure of the petals, the drupaceous fruit, and the curved embryo with inferior radicle, and thid affinity is prooably"a distant one. * Xinn.Tr. xil 355. § In Hook. Journ. Bot. iv. 7$. t Roxb. rl Ind. ed. Wall ii 308. |[ Mus. LnwL Bat. i. 808. t. 44. t Bijdr. p. 28. ^ Lindley'i Veg. Kingd. 3rd ed. p.